We present the results from a series of experimental investigations into the hydrodynamic instabilities that occur in radiative blast waves. In particular, we examine the Vishniac instability in which the perturbation modes oscillate in time and, for certain mode numbers and polytropic index of the medium, can exhibit a growth in their amplitudes. Experiments were conducted on the GHOST laser laboratory in which a source of atomic clusters was irradiated by a 1J-2J, 115fs laser pulse to produce cylindrical blast waves. The thrust of this thesis falls into two categories. First, we analyze the effects radiative cooling has on the evolution of blast waves such as the lowering of the effective polytropic index and consequently the lowering of their deceleration parameter. Radiation from the blast wave surface results in a preheated ionization precursor in the upstream material and is indicated by a gradual decline in the electron density profile of the blast wave rather than a sharp jump. This mechanism, if strong enough, can also create a secondary shock wave to form ahead of the main blast wave. The second set of experiments investigates the temporal evolution of longitudinal perturbations induced on the blast waves by use of a transverse interferometric beam that modifies the cluster medium prior to the onset of the main pump beam. These perturbations are analyzed and compared to theory set forth in Vishniac's mechanism for oscillatory instabilities and their growth rate. / text

The near field blast–wave propagation dynamics has been a subject of intense research in recent past. Since experiments on a large scale are difficult to carry out, focus has been directed towards recreating these blast waves inside the laboratory by expending minuscule amounts of energy(few joules),which have been termed here as micro–blast waves. In the present study, micro-blast waves are generated from the open end of a small diameter polymer tube (Inner Diameter of 1.3 mm)coated on its inner side with negligible amounts of HMX explosive (~18 mg/m), along with traces of aluminium powder. Experimental, numerical, and analytical approaches have been adopted in this investigation to understand the generation and subsequent propagation of these micro–blast waves in the open domain. Time–resolved schlieren flow visualization experiments, using a high speed digital camera, and dynamic pressure measurements (head–on and side–on pressures) have been carried out. Quasi one dimensional numerical modeling of the detonation process inside the tube, has been carried out by considering the reaction kinetics of a single(HMX) reaction to account for the reaction dynamics of HMX. The one dimensional numerical model is then coupled to a commercial Navier– Stokes equation solver to understand the propagation of the blast wave from the open end of the tube. A theory that is valid for large scale explosions of intermediate strength was then used for the first time to understand the propagation dynamics of these micro–blast waves. From the experiments, the trajectory of the blast wave was mapped, and its initial Mach number was found to be about 3.7. The side–on overpressure was found to be 5.5 psi at a distance of 20 mm from the tube, along an axis, offset by 30 mm from the tube axis. These values were found to compare quite well with the numerically obtained data in the open domain. From the numerical model of the tube, the energy in the blast wave was inferred to be 1.5 J. This value was then used in the analytical theory and excellent correlation was obtained, suggesting the exciting possibility of using such theories, validated for large-scale explosions, to describe these micro–blasts. Considering the uncertainties in the approximate model, a better estimate of energy was obtained by working back the energy(using the analytical model) from the trajectory data as 1.25 J. The average TNT equivalent, a measure of its strength relative to a TNT explosion, was found to be 0.3. A few benchmark experiments, demonstrating the capability of this novel blast device have also been done by comparing them against the extant large–scale explosion database, suggesting the possibility of using these micro–blast waves to study certain aspects of large–scale explosions.

No description available.

Concrete is a widely studied material with a composite nature. It is used both in civil and military buildings and infrastructures. An issue of great importance is the protection of people from terrorist attacks that target critical infrastructure. Explosions, detonations and/or projectile impacts are some of the most severe actions a concrete structure can face. Experimental analysis is necessary in order to understand and predict the response of a structure to such dynamic and strain rate sensitive conditions. However, as the cost of performing experiments is significant and numerical simulations offer improved blast and impact analysis capabilities, there is an effort to limit experiments to validation purposes. In recent years, many researchers have studied the impact loads transferred to reinforced concrete (RC) structures both through direct projectile impacts or blast waves at both near and far field. The aim of the current study is twofold. First, to investigate contact detonations on this type of material (RC), since literature can provide us with limited information. Secondly, to assess the behaviour of the RC structure under combined ballistic impact and contact detonation of a very specific geometry of projectile (HESH) that exists currently on the market and behaves differently from the normal projectiles that consist of one single material. The author analysed and discussed in depth the response of RC members exposed to contact detonations. More precisely, the effect of the mass of explosive (C4) on pressures, impulses and energy balances. Also, she investigated the kinematic response of RC slabs and the structural role of the reinforcing bars. The driving force of this RC structures. Currently, the majority of studies regarding contact blast are focusing either on innovative types of concrete or normal concrete. However, normal concrete is investigated as a control parameter (to prove the effective resistance of the innovative material) rather than a detailed study on the behaviour of the material. Thereafter, the author analysed the response of a RC wall under the combined effect of kinetic energy (terminal ballistics) and contact detonation caused by the impact of a 90 mm HESH (High Explosive Squash Head) projectile fired from a distance of 70 m. The aim was to investigate the response of the structural member under the superposition of those two actions and analyse the combined effects of the impact velocity and detonation on the response of the structure. The numerical modelling is based on a Multi-Material-Arbitrary-Lagrangian-Eulerian approach (MMALE, using LS-DYNA) using the Winfrith concrete constitutive material model to investigate the dynamic response of the RC members under high strain rate conditions. The efficiency of the proposed numerical modelling is validated with experimental results - based on open-arena testing - and provided by the Royal Military Academy of Belgium. Some of the key findings of this research are that the increase of the amount of the explosive affects the damage failure of the RC members from flexural failure to shear failure. In addition, fitting curves that could be used in design, were proposed, that show the relation between the mass of explosive and the resulting pressures and impulses, within the tested range. In the case of the combined blast and impact scenario, the detonation was found to dominate the structural response of the RC slab.

In many parts of the world, civilians and peacekeepers are exposed to potentially serious injury from blasts and explosions. Providing insight into the trauma thresholds for blast injury is necessary for the development of blast protection equipment and identification and subsequent treatment of blast injury. [Phillips, 1988] Blast injury can be categorized as primary, secondary, tertiary, quaternary and quinernary, corresponding to different aspects of the blast loading and injury mechanisms. Primary blast injury occurring in the lungs is of importance, since lung injury results in one of the highest rate of blast mortality. Much of the existing blast injury data was obtained from animal testing with sheep and subsequently extrapolated to humans using scaling techniques. More recently, mathematical, experimental and numerical models have been developed and employed to investigate blast injury. In this study, a detailed finite element model of a sheep thorax and human thorax (developed at the University of Waterloo) was used to predict primary blast lung injury based on a range of blast loading conditions. The models were developed based on available anatomical data and material properties to model the organs and tissues, and were evaluated using the LS-Dyna explicit finite element code. The models were previously validated for the prediction of lung PBI using Friedlander-type blast waves. All results were compared to existing literature to further verify and validate the numerical models as wells as to provide insight on the effect of loading conditions on blast injury. The blast loading input for these simulations used idealized blast waves, based on a blast physics approach. Blast loads were verified using the Chinook CFD software. The effects of idealized blast waves on predicted lung injury were investigated to determine the importance of peak pressure, blast wave duration and impulse. The duration and peak pressures for the waves were selected based on the Bowen and UVa curves, and included a right angle triangular shape and a square wave to allow for the different parameters to be considered. These results were compared to the Bowen and revised Bowen injury models. The results show that the peak overpressure is dominant in predicting injury for blast loads with long durations (>8 ms). The impulse was dominant in predicting injury for blast loads with short durations (<1 ms). For blasts loads with intermediate durations (1 ms < 8 ms) both the shape of the blast load wave and peak overpressure play a role in primary blast lung injury. The effect of orientation of the body position on primary blast lung injury was investigated. Simulations were performed using the sheep and human numerical models along with a model of a commonly used experimental device, the Blast Test Device (BTD) cylinder. These models were oriented in different positions by rotating the body relative to the blast flow. Injury results for the BTD were calculated using the Injury 8.1 injury prediction software. The BTD simulations served several purposes; it was used as a reference for the human and sheep simulations and its effectiveness as a tool to predict body orientation was evaluated. In general, all of the models predicted appropriate and similar levels of injury for the body in its default orientation, and these predictions were comparable to the accepted injury levels for this insult. For other orientations the BTD was not able to predict the appropriate blast injury. This highlighted the importance of proper placement and orientation of the BTD when used in simulations or physical experiments. The overall injury (based on the results from the right and left lung) predicted by the sheep and human thorax was similar for all orientations. However, very different results were obtained when the predicted injury for the right and left lungs was compared. The differences between the sheep and the human were examined and the differences in injury between the right and left lung is a result of the differences in anatomy between the two species. This study has evaluated the importance of blast wave parameters in predicting primary blast injury, an important consideration for the improvement of blast protection, and the effect of body orientation on primary blast injury, an important consideration for experimental testing and a starting point for the evaluation of complex blast loading. Future work will focus on the evaluation of injury in complex blast environments.

In many parts of the world, civilians and peacekeepers are exposed to potentially serious injury from blasts and explosions. Providing insight into the trauma thresholds for blast injury is necessary for the development of blast protection equipment and identification and subsequent treatment of blast injury. [Phillips, 1988] Blast injury can be categorized as primary, secondary, tertiary, quaternary and quinernary, corresponding to different aspects of the blast loading and injury mechanisms. Primary blast injury occurring in the lungs is of importance, since lung injury results in one of the highest rate of blast mortality. Much of the existing blast injury data was obtained from animal testing with sheep and subsequently extrapolated to humans using scaling techniques. More recently, mathematical, experimental and numerical models have been developed and employed to investigate blast injury. In this study, a detailed finite element model of a sheep thorax and human thorax (developed at the University of Waterloo) was used to predict primary blast lung injury based on a range of blast loading conditions. The models were developed based on available anatomical data and material properties to model the organs and tissues, and were evaluated using the LS-Dyna explicit finite element code. The models were previously validated for the prediction of lung PBI using Friedlander-type blast waves. All results were compared to existing literature to further verify and validate the numerical models as wells as to provide insight on the effect of loading conditions on blast injury. The blast loading input for these simulations used idealized blast waves, based on a blast physics approach. Blast loads were verified using the Chinook CFD software. The effects of idealized blast waves on predicted lung injury were investigated to determine the importance of peak pressure, blast wave duration and impulse. The duration and peak pressures for the waves were selected based on the Bowen and UVa curves, and included a right angle triangular shape and a square wave to allow for the different parameters to be considered. These results were compared to the Bowen and revised Bowen injury models. The results show that the peak overpressure is dominant in predicting injury for blast loads with long durations (>8 ms). The impulse was dominant in predicting injury for blast loads with short durations (<1 ms). For blasts loads with intermediate durations (1 ms < 8 ms) both the shape of the blast load wave and peak overpressure play a role in primary blast lung injury. The effect of orientation of the body position on primary blast lung injury was investigated. Simulations were performed using the sheep and human numerical models along with a model of a commonly used experimental device, the Blast Test Device (BTD) cylinder. These models were oriented in different positions by rotating the body relative to the blast flow. Injury results for the BTD were calculated using the Injury 8.1 injury prediction software. The BTD simulations served several purposes; it was used as a reference for the human and sheep simulations and its effectiveness as a tool to predict body orientation was evaluated. In general, all of the models predicted appropriate and similar levels of injury for the body in its default orientation, and these predictions were comparable to the accepted injury levels for this insult. For other orientations the BTD was not able to predict the appropriate blast injury. This highlighted the importance of proper placement and orientation of the BTD when used in simulations or physical experiments. The overall injury (based on the results from the right and left lung) predicted by the sheep and human thorax was similar for all orientations. However, very different results were obtained when the predicted injury for the right and left lungs was compared. The differences between the sheep and the human were examined and the differences in injury between the right and left lung is a result of the differences in anatomy between the two species. This study has evaluated the importance of blast wave parameters in predicting primary blast injury, an important consideration for the improvement of blast protection, and the effect of body orientation on primary blast injury, an important consideration for experimental testing and a starting point for the evaluation of complex blast loading. Future work will focus on the evaluation of injury in complex blast environments.

La formation et l'activité des objets compacts, tels que Trous Noirs ou étoiles à Neutrons, s'accompagne couramment d'éjection de matière ionisée sous forme de jets à la vitesse proche de celle de la lumière (vitesses relativistes). Se propageant dans le milieu environnant, par exemple Milieu Interstellaire, ces jets conduisent inéluctablement à la formation d'ondes de choc relativistes. Une forte turbulence magnétique et une population d'électrons accélérés sont requises afin de tenir compte de l'émission radiative non-thermique de ces chocs. L'approche naturelle de ce problème, décrivant de manière auto-consistante la structure du choc non-collisionnel, est celle de la physique cinétique des plasmas en régime relativiste. L'aspect essentiel de cette approche est l'étude du précurseur du choc, sous forme d'un faisceau de protons très énergétiques. Un ensemble d'instabilites plasma y prend lieu et dissipe l'energie du choc sous forme de micro-turbulence électromagnétique, électrons chauffés et particules accélérées. Ce cadre conceptuel emmène à reconsidérer le processus de transport de particules charges autour du choc. Deux études indépendantes, effectuées pendant la thèse, montrent que les lois de diffusion en aval et amont du choc se mettent sous une forme concise, en loi de puissance en fonction de l'énergie des particules et de l'intensité de la micro-turbulence magnétique. Les lois de diffusion, dérivées à l'aide des simulations Monte-Carlo et analytiquement, chiffrent l'énergie maximale des protons accélérés au choc à 10^15 eV, si le facteur de Lorentz du choc est très grand devant 1. Cette limite se situe loin de l'énergie maximale des Rayons Cosmique et rend les chocs relativistes comme accélérateurs de particules inefficaces aux énergies les plus extrêmes. Le rayonnement, issu de l'accélération des électrons, atteint plusieurs GeV et corrobore l'idée que les chocs externes des Sursauts Gamma peuvent émettre à de telles énergies. L'approche alternative de l'étude des chocs, simulations Particle-In-Cell, m'as permis d'étudier la formation, structuration et évolution des chocs modérément relativistes dans une géométrie spatiale 1D. L'auto-reformation du front d'un choc perpendiculaire, connue dans le régime non-relativiste, persiste dans le régime moyennement relativiste et exhibe un front de choc non-stationnaire. A magnétisation basse, les électrons sont préchauffés dans le pied du choc par l'instabilité de Buneman entre protons réfléchis et électrons incidents, mais leur température en aval du choc reste plus faible que celle des protons. A magnétisation croissante, l'instabilité Maser Synchrotron devient essentielle dans la structuration du front de choc, avec émission d'un fort précurseur électromagnétique a partir du front de choc. Dans ce cas les électrons se mettent en équipartition avec les protons. Ces simulations 1D ne montrent pas d'évidence d'accélération des particules et des simulations 2D (3D) sont nécessaires. / The formation and activity of compact objects such as Black Holes and Neutron Stars results in the ejection of ionized matter in the form of jets with velocities close to $c$ (relativistic). The interaction of such powerful jets with the external medium forms shocks, eventually relativistic.A strong self-g???enerated magnetic micto-turbulence and a population of accelerated electron are required to explain the observed non-thermal radiation of these shocks. A natural approach to the study of the structure of a non-collisionnal shock involves kinetic treatement of plasma processes in the relativistic limit. This approach is adopted in the present thesis.Consequently, charged particle transport laws need to be studied carefully taking to acount self-consistent magnetic micro-turbulence at the shock. Two different studies of particle transport at each side of the shock (downstream and upstream) show that the diffusion laws take a concise form as a power law in energy ($D \propto E^2$) and the micro-turbulence strength. Both Monte-Carlo simulations and analytic studies are in agreement and, if the shock Lorentz factor is much greater than 1, it is found that the maximum energy of accelerated protons is $10^{15}$eV. A physical mechanism is also provided to explain how electrons attain the equipartition with protons at the shock. Finally, the radiation from accelerated electrons at the shock can reach several GeV in a synchrotron-like spectrum.In the second part of the thesis, I used 1D3V PIC simulations to study mildly relativistic shocks structure and their time evolution. The prependicular shock front self-reformation, well-known in non-relativistic limit, persists at mildly relativistic speeds. At low magnetization ($\sigma \ll 10^{-2}$), electrons are pre-heated in the shock precursor by the Buneman instability between reflected ions and incident electrons. At higher magnetizations ions form a coherent cyclotron loop at the front and the Maser Synchrotron Instability is essential for the shock structure by emitting a strong electromagnetic precursor, responsible for electrons heating up to equipartition with protons. No particle acceleration is seen in these 1D3V simulations.